It happens every few years: a star-studded college team, fresh off a dominating performance, is compared favorably to the worst squad in the pros. A few years ago, it was the Anthony Davis-led Kentucky powerhouse vs. a pitiful Charlotte Bobcats outfit that set a new NBA low for winning percentage in a season. Now, it’s Kentucky’s 2014-15 team, which trounced fifth-ranked Kansas on Tuesday, vs. the tank-tactic Philadelphia 76ers.
In a radio interview this week, former Kentucky guard Eric Bledsoe (now of the Phoenix Suns) said his alma mater would own the hypothetical matchup. “I’m definitely taking Kentucky,” Bledsoe said when asked who would win a best-of-seven series between the Wildcats and Sixers. “I think Philly would probably get maybe one game.”
As others have pointed out, that’s absurd. As awful as they are — and boy, are they awful — the Sixers’ roster represents an All-Star team of college players from the past few seasons:
- Center Nerlens Noel was first-team All-SEC two seasons ago; if he were in college now, he’d likely be considered the best player in the country;
- Point guard Michael Carter-Williams was an honorable mention All-American as an NCAA sophomore; had he not left Syracuse, he would likely be considered the best guard in the college game;
- Guard Tony Wroten made first-team All-Pac 12 as a freshman two years ago;
- And rookie K.J. McDaniels was named to the All-ACC first team last season.
These are all players who could still have NCAA eligibility if they had they not departed early for the NBA.
To put these anecdotes in a more rigid framework, we can return to the same projection methodology we used for our NBA preview. The premise there was to take Real Plus/Minus data from last season, adjust for aging effects and regress to the mean to estimate each player’s true talent level. In this case, we also need to account for any information gleaned from the first three weeks or so of the NBA season; since RPM isn’t available for 2014-15 yet, I’m using a weighted average between a player’s current 2014-15 Statistical Plus/Minus (a box score-based statistic designed to emulate RPM in situations where RPM isn’t available) and our projected RPM ratings from the preseason.
By this accounting, the 76ers are made up exclusively of below-average players. In fact, 54 percent of the team’s minutes have gone to players below the replacement-level RPM threshold of -2 points per 100 possessions. If we compute a minutes-weighted aggregation of ratings for Philadelphia’s 2014-15 roster (and adjust for the fact that the Sixers are usually trailing in their games), we’d expect them to lose to an average NBA team by 8.9 points per 100 possessions, which is ridiculously bad for a team’s true talent level. (Other teams have posted worse seasons, but that’s usually because they have bad rosters that perform below their talent levels.)
But as bad as the Sixers are, Kentucky’s roster would translate to an even worse team at the NBA level. Looking at how well players’ draft slots predicted their rookie RPM performances (and, in turn, how well ultra-early mock draft rankings like these predict a player’s draft slot), and turning the clock back on those rookie-year RPM projections using an aging curve, we can also estimate an NBA-equivalent 2014-15 RPM talent level for every player on the Wildcats’ roster. Their top prospect, Karl-Anthony Towns (who sits at No. 3 in ESPN’s class-of-2015 draft rankings), would translate to a -2.6 RPM performance at the NBA level this season, worse than almost every player on Philadelphia’s roster. Every other Kentucky player, from Willie Cauley-Stein to Alex Poythress and the Harrison twins, grade out with various shades of putrid RPM ratings that would embarrass even Hakim Warrick.
The aggregation of Kentucky’s NBA-translated RPM scores would predict an efficiency margin of -13.6 against an average NBA team. Even after accounting for the fundamental uncertainty surrounding projection models, such a team would lose about 88 of every 100 games against average NBA competition on a neutral court. The Sixers are far from average, of course, but even they would be expected to beat the Wildcats 74 percent of the time at home, and 56 percent of the time on the road. Using those probabilities to simulate Bledsoe’s hypothetical best-of-seven matchup, Kentucky would lose the series 78 percent of the time even if we gave them home-court advantage, most frequently falling in six games.
It’s also likely those numbers vastly overestimate Kentucky’s chances. The NBA doesn’t let prospects who are not believed to be pro-caliber players take the court in games, so we don’t really know what kind of performance expectations to set for those Wildcat players who will never play in the NBA. This means the aforementioned translations are naturally biased toward inflating Kentucky’s rating. Furthermore, it’s probably not appropriate to assume the same uncertainty level around Kentucky’s rating that we would for an NBA team. Ironically, because we know they are a college team with no historical track record against NBA teams, we should probably be more certain that their horrible aggregated RPM rating is appropriate, because the sample of (NCAA) teams they’re being drawn from is known to be inferior to the NBA.
Kentucky would not be favored against any NBA team, even one as horrid as the Sixers have been. The Wildcats look good when dismantling collegiate opponents, but the NBA is, quite literally, a whole different ball game.